Infrared and visible light image fusion method based on generative confrontation network

Abstract

The invention discloses an infrared and visible light image fusion method based on a generative confrontation network. The network realizes automatic learning of fusion features and fusion rules in the process of training the convolutional neural network; infrared images and visible light images are simultaneously input to the generator in the generation confrontation network, and the generator performs image fusion operations and outputs the final fusion image. The present invention breaks through the limitations of manual design of fusion features and rules, and uses the generator in the generation confrontation network to automatically learn fusion features and rules, and the fusion image generated by the present invention can well maintain the basic radiation information of infrared images, and has Rich texture information, while the fusion image has good visual effect and high target salience.

Description

technical field [0001] The present invention relates to the technical field of infrared image fusion, in particular, the present invention relates to a technical proposal for fusion of infrared and visible light images based on a generative confrontation network. Background technique [0002] The infrared imaging system has strong anti-interference ability, good concealment performance, strong atmospheric penetration ability, obvious target, and can adapt to many special occasions. However, due to the characteristics of the infrared detector itself, such as high sensitivity and large dynamic range, as well as various noise interference in the complex working environment, the infrared image shows the characteristics of high background and low contrast. The specific performance is that the imaging scene only accounts for a small part of the dynamic range of the entire infrared imaging system, and the image contrast is poor and blurred. On the contrary, the visible light image...

AI Technical Summary

Problems solved by technology

However, these methods all use the same strategy to select the same salient features in the infrared image and the visible light image for fusion, so that the target that was originally salient in the infrared image is no longer salient in the visible light image.

At the same time, these methods have to manually design fusion features and fusion rules, making it difficult to make new breakthroughs in the field of infrared and visible light images.

Images